waltje a écrit : ↑27 avr. 2019 21:36
I have found the meaning of the two switch blocks, and some others. See page 239 and on in the book I posted here:
http://waltje.varcem.com/book.pdf. That is literally the
ONLY info I could find on it...
Thank you much, Fred! Best 2.000 page book I've ever read
Despite some of my research efforts were basically correct, I didn't even suppose these functions, they're almost impossible to guess.
As i've understood "OFF/ON" in that table shows default value, but i don't know what means "*"-symbol.
I've compared that with my schemes and rechecked the board. I have some corrections and questions.
Question mark means I am not sure which position (ON/OFF) does the specified action.
First, SW2:
Code : Tout sélectionner
#1 ON? = mouse enable
#2 ON = give IRQ3 to mouse (IRQ5 in your manual is wrong)
#3 ON = give IRQ5 to mouse (IRQ3 in your manual is wrong)
#4 ON?? = enables CP8, Serial, Parallel (otherwise i don't understand this function)
#5 ON = set SCSI address 328-32Fh, OFF = 328-32Fh
#6 ON = give IRQ7 to SCSI
#7 ON = give IRQ15 to SCSI
#8 ON = give IRQ5 to Parallel port (=COM2)
#9 ON = give IRQ7 to Parallel port (=COM2)
#10 ON = give IRQ3 to CP8 (=COM1)
Seems that available serial port on this machine should be COM0 with fixed IRQ (4?), CP8 interface implemented via COM1 and COM2 is parallel port.
So theoretically, if one needs, he can continue reverse-engineering and get one more COM port from 20-pin CP8 interface.
Then, SW1. Howled in relief that half of switches ( #3, #6-8) does nothing. The rest are:
Code : Tout sélectionner
#1 = memory setup
#2 = memory setup
#4 ON = sets Mono mode for external (what else means addon?) ISA videocard
#5 ON = enables external videocard
To conclude, this PC already has not so much switches but most of them are just to manage built-in devices.
Nothing close to any normal 286 PC, no hardware turbo key and strange memory setup.
This is my latest version of switch setup.
I hope Lenord will be able to "map" the switches to the bits in the keyboard controller. The BIOS has to
be able to read them, and this usually happens through the keyboard controller's "input port" bits.
For now, my emulated M45 works with Lenord's V2.0 BIOS, but, for some reason, not with Fabien's
I've found three - J1 jumper (32-pin, port 15), switch SW1 #4 (33-pin, port 16) and Turbo LED (24-pin, port 23).
Other switches aren't connected with keyboard controller directly. I will try to find something else but it's almost impossible because they may come through lots of logics, some of which is pre-programmed. I've tried that for SW2, maybe that helps.
It's hard to be emulated because we even don't have any datasheet for its chipset(
I can't imagine how you got that working, but that's great!
I do not really understand the corresponding part for the memory. I'll have to re-try.
Me too. First two settings do compare with my results.
Third setting isn't clear, but something connected with second bank.
Fourth setting (=mine) means nothing useful and there must be misprint (double OFF OFF)
I hope now you could check that.
Please try with 2x, 4x and 6x 1Mb simms in slots and fill-in 3x tables in the same order.
I will do the same but I am not sure that my board works correctly.
Also if you have Inport mouse(or at least COM with DB25 adapter), you could check if this mouse stops working when you change #1 or #4 in SW2 block. I don't know, but there should be any DOS software which can be used to test mouse.
I came to an idea, that my board is ok but was cropped to 8MHz and 2Mb RAM by manufacturer.
I think, that our boards have different pre-programmed logics, especially Z322 which is not soldered thus replaceable.
I've restored clock scheme from the CPU:
You see, that Z322 controls both memory switches and clock generation. I couldn't find chipset's datasheet but anyway, this could be the reason.
So I need more photos of some parts of your board to ensure. I have to check the labels of the chips, so i need closer look.
I've marked them with red rectangles.